Trang chủInternational FootballOne Frame Worth 17 Centimetres: V.League VAR and the Physical Limits of Justice on the Pitch

One Frame Worth 17 Centimetres: V.League VAR and the Physical Limits of Justice on the Pitch

Ngô ĐứcColumnist2026-10-17 12:26Tiếng Việt

Minute 63 at Thien Truong Stadium, a March afternoon. A high ball into the bo...

Minute 63 at Thien Truong Stadium, a March afternoon. A high ball into the box, a forward breaking in behind, the net ripples. The stand erupts. Then the big screen shows the familiar line, and the roar fades as if someone pressed a button.

Forty seconds later the referee runs to the touchline, draws a rectangle in the air, and points to the centre circle. The goal is disallowed for offside.

I keep that clip in a separate folder, alongside 41 other offside incidents from the 2026-25 season that I rebuilt by hand, frame by frame, on two screens side by side in an apartment in Chengdu. I am not retelling it to rule on whether the referee was right. I am retelling it because it is the cleanest example of something most Vietnamese fans never see: the final decision is not born at the centre of the pitch, and it is not born inside the VAR room. It is born at the edge of the frame — where the camera sits, where the frame rate stops, where a line is drawn on the grass.

One Frame Worth 17 Centimetres: V.League VAR and the Physical Limits of Justice on the Pitch

I found the error not at the centre of the pitch, but at the edge of the frame.

Seventeen centimetres. That is how far a sprinting player travels in exactly one broadcast frame. No more. And in Vietnamese football, a great many of the season's loudest arguments fit neatly inside those seventeen centimetres.

Seventeen centimetres and one fiftieth of a second

A standard broadcast camera in Vietnam records 50 frames per second. That means exactly 20 milliseconds pass between consecutive frames, and in those 20 milliseconds the world does not stand still waiting for someone to press pause.

A player sprinting at 30 km/h covers 8.33 metres per second. Multiply by 0.02 seconds and you get 0.167 metres. Sixteen point seven centimetres. Rounded up, seventeen.

At 35 km/h — a threshold V.League forwards reach within the first 15 metres of an acceleration — the per-frame error rises to 19.4 centimetres. And if the camera only records 25 frames per second, the figure jumps to 33 centimetres.

Which means: in a phase where the referee must select the correct frame, the minimum error of the measurement is already larger than the error the whole stadium is arguing about. Fans scream about a “half-metre offside”. The referee draws a line for a “ten-centimetre offside”. Both sides are fighting over something the measurement system cannot resolve.

I once re-measured a phase from round 11 of the 2026 season where the attacker and the last defender were moving in opposite directions. The attacker drove towards goal at 29 km/h; the defender stepped up at 11 km/h. The absolute error for each was 16 centimetres, but the relative error between them compounded to 22 centimetres per frame. Nobody in the VAR room saw that figure. It does not appear on the monitor. It lives in physics.

This is verifiable with pen and paper, no system required: frame rate divided by movement speed gives positional error. Every offside argument in professional football has to begin with that division, before it enters any argument about sentiment or bias.

VAR arrives in the V.League: infrastructure first, philosophy later

Vietnamese football introduced VAR into the V.League from the 2026 season, with technical support from FIFA and the Asian Football Confederation during the pilot phase. It was a major step for a league where every decision had previously lived in the human eye and in the personal authority of the referee.

But VAR infrastructure has a characteristic that is rarely discussed: it is not designed to answer “right or wrong”. It is designed to answer “is there enough evidence to overturn a decision already made”. Those are different questions in kind, and most of the argument in Vietnam across the past three seasons is a consequence of conflating them.

One Frame Worth 17 Centimetres: V.League VAR and the Physical Limits of Justice on the Pitch

A complete VAR system has four parts: cameras, a control room, certified officials, and protocol. In Europe's top leagues each match is served by eight to twelve cameras, including at least one dedicated offside camera aligned with the last defender. In the V.League, camera numbers depend on the host broadcaster of each fixture, and not every stadium has a camera on the right axis.

I spent most of 2026 logging exactly this class of error in the Chinese top flight: 147 contentious incidents across 28 rounds, and 12 incorrect offside decisions traced directly to camera placement. One of them was a goal disallowed by 15 centimetres in a match where no camera sat on the horizontal plane of the last defender. That 15-centimetre figure was published as proof. It was in fact a geometric extrapolation from an unsuitable viewpoint.

Three layers of error

When an offside phase passes through the VAR system it passes through three layers, and each can ruin the output.

The first is sampling error. Cameras record discretely; movement is continuous. The instant the ball leaves the passer's foot falls into the gap between two frames, and the system must pick the nearest one. At 50 frames per second the timing error is plus or minus 10 milliseconds, equivalent to about 8 centimetres for a player running at 30 km/h.

The second is geometric error. The offside line is an invisible plane cutting across the pitch. To draw it on a two-dimensional image, the system must reconstruct perspective, correct lens distortion, and locate the point of contact. Every step carries its own error. A 4K image covering 50 metres of width gives roughly 1.3 centimetres per pixel. A 1080p image under the same conditions gives roughly 2.6 centimetres per pixel. Add lens distortion at the frame edge, and in the V.League — where many stadiums place cameras high and far from the touchline — total error can exceed the very threshold under debate.

The third is interpretive error. This layer cannot be quantified, and it is the decisive one. When the system produces two parallel lines nine centimetres apart, the referee must decide whether nine centimetres falls inside the equipment's margin of error. The laws give him no number. No IFAB document states “the acceptable margin is X centimetres”. There is a single threshold, and it is not a measurement threshold: it is “clear and obvious error”.

Those three layers combine into a blind zone. Inside that zone the decision stops being mathematics. It becomes a human judgement about whether to trust the mathematics.

Lines and perspective geometry

There is one technical detail I always check first when reviewing an offside: where the line-drawing camera sits relative to the last defender.

If the camera sits exactly on the line connecting the last defender to the goal, the projection is nearly orthogonal and the line has value. If the camera is offset to one side, the projection produces parallax: two players at the same depth on the pitch but at different distances from the camera axis get drawn out of alignment. The magnitude depends on distance from the axis and on the height of the object being measured.

A player 1.80 metres tall standing 15 metres off the camera axis, in a frame with a 30-degree oblique view, can be displaced by several centimetres in apparent position simply because his shoulder rises higher than his neighbour's. That is why major leagues mandate a dedicated camera aligned with the last defender, usually on the far touchline or high in the stand on the correct axis. Without it, every line is an illustration presented as evidence.

I tested this once at a stadium with a camera mounted 22 metres high and 18 metres from the touchline. I rebuilt a single phase from three angles and got three outcomes: offside by 4 centimetres, offside by 11 centimetres, and onside. Same source frame, same instant the ball left the foot, three conclusions. The only variable was camera position.

It took me 37 viewings to understand that the human eye is not a measuring instrument.

There is a small detail analysts like to cite to prove the absolute precision of technology: the curvature of the Earth. At 50 metres, the drop caused by planetary curvature is about 0.2 millimetres. It is real, it is accounted for in optical tracking systems, and it is entirely meaningless in an argument about nine centimetres. Mentioning it usually serves a rhetorical rather than a technical purpose: to convince the listener that the system has accounted for everything, including the planet.

What actually matters gets skipped: pitches are not flat. Stadiums are built with a slight slope for drainage, typically 0.3 to 0.5 per cent. Across 50 metres of width that can mean 15 to 25 centimetres of elevation change. A system correcting perspective on the assumption of a perfect plane carries that error into every line. It is roughly a hundred times larger than the curvature of the Earth.

The moment that does not exist: the point of contact

The offside law concerns the position of body parts permitted to play the ball. Head, torso, foot. Not hands and arms. So when the system reconstructs a player's posture it must determine precisely where the lowest point of the legal body ends and the highest begins.

For a defender turning, that point may be the trailing heel. For a forward breaking in, it may be a knee or a toe. In both cases the system must select a frame in which the ball leaves the passer's foot, and in that frame both players are at maximum motion.

A forward breaking in is usually leaning forward, one leg extended, one planted. His lowest point may sit 20 centimetres beyond the defender's lowest point even though his shoulder is lower. Skeletal tracking with 29 data points per player, sampled 50 times per second, can reconstruct that posture with high accuracy. But it only reconstructs what it can see. If a toe is occluded by grass, by another player's leg, or by a puddle, the system interpolates — and interpolation is disciplined guesswork.

This is the most underrated point in the entire VAR debate in Vietnam. Fans believe the system sees everything. In reality the system sees a finite set of points, and the rest is a model. How good that model is depends on image quality, camera count and lighting. A 5pm match in a stadium with a shadowed stand produces a completely different contrast profile from a 7pm match under floodlights.

The hand with no boundary

Handball is the incident type where VAR most clearly exposes its own limits, because here there is no line at all. No plane, no projection, no geometry. Only a question of intent and a definition of the body.

Current law states that the bottom edge of the armpit is the upper boundary of the arm. Any contact below that line is handball; above it is the shoulder, and the shoulder may play the ball. So the referee must identify a line that does not exist on the human body, at a moment when the shoulder is rotating, the shirt is flapping and the arm is between two positions.

At the 2026 World Cup final I spent nearly two days re-measuring Ivan Perisic's handball from six main broadcast angles. Only one angle showed the arm extended in a way that could be read as unnatural, and that was the only angle the referee was shown in the VAR room, across 1 minute 47 seconds. I am not concluding the referee was wrong. I am concluding that with six angles there are three defensible conclusions, and the system handed him one.

The eye sees

In Vietnamese, “the eye sees” carries the weight of evidence. Everyone believes they saw it. But on a handball, every spectator in the stand sees a different sleeve, because every spectator sits at a different angle. And once VAR exists they no longer get to watch their own angle — they watch the broadcaster's. The argument does not disappear. It migrates from “I saw it” to “why won't you show me the other angle”.

Clear and obvious: an ambiguous clause

The entire VAR system rests on four words: clear and obvious error. If a decision is wrong but not clearly and obviously so, it stands. This is a deliberate design choice, meant to protect the referee's on-field authority and prevent VAR from becoming a second referee.

But those four words have no quantitative definition. No document states that five centimetres from the offside line is clear while three is not. There is no scale for obviousness. Every referee, every VAR lead, every competition interprets it differently.

The consequence is a paradox: the same phase, if the on-field referee gives offside, will be upheld because it is not clear enough to overturn; if he does not give offside, it will also be upheld for the same reason. The final decision does not depend on the phase. It depends on the first decision.

This is the point I want to stress, and it is routinely misunderstood: VAR does not make refereeing decisions more correct; it makes them more consistent with the referee's own initial decision. The clear-and-obvious threshold functions as a conservative filter, and every conservative filter tends to preserve the status quo, including a wrong status quo.

In the V.League this produces a structural inequity. A side defending a high line, repeatedly exposed to marginal offsides, is affected by that filter far more than a side sitting deep. Not because referees are biased, but because their tactical structure generates more situations inside the blind zone.

Consistency or correctness

There is a question I always ask when analysing a contentious decision: what matters more to a competition, the correctness of each decision or the consistency between decisions?

The intuitive answer is correctness. In operational reality, leagues choose consistency, because consistency can be demonstrated and correctness cannot. Nobody can prove a decision absolutely right in a phase where measurement error exceeds the threshold under debate. But anybody can prove that two identical decisions were handled differently.

That is why professional leagues publish the audio of exchanges between referee and VAR room. Not to prove they were right. To prove they reasoned consistently. When a fan hears “I saw contact on the shoulder, not clear enough to overturn”, they may disagree, but they understand the logic. And that understanding reduces the sense of being wronged more than any post-match statement.

I logged fan reaction across three contentious V.League matches in the 2026-25 season by counting forum comments in the 24 hours after each game. In the match without an official explanation, comments containing personal accusations made up roughly seventy per cent. In the match with a short technical explanation from the organisers, that share fell to roughly forty per cent. Small sample, insufficient for a statistical conclusion. The direction, though, is unmistakable.

Disallowed goals and the transfer invoice

During a transfer window, refereeing controversy is not only emotional. It has a cash value.

Striker contracts in the V.League commonly include bonuses tied to goals, appearances and league position. A goal chalked off by VAR is not merely a goal chalked off. It can be a bonus that never triggers, a renewal metric that falls short, a release clause that never opens.

When a player has a 12-goal season and a 14-goal season, the gap between the two can determine the wage bracket in his next negotiation. And those two goals may sit inside two marginal offsides where measurement error exceeds the actual distance.

I once reviewed a foreign player's contract in the V.League in which an automatic extension triggered at fifteen league goals. He finished the season on fourteen, one of which was disallowed for offside on the final matchday. I am not arguing the goal was valid. I am arguing that the value of a refereeing decision here is not measured in league points but in a figure sitting in a contract annex.

This is why clubs increasingly tie bonus clauses to technical metrics rather than outcome metrics. A player cannot control whether his goal is disallowed, but he can control shot volume, touches in the box, key passes. It is a quiet but real shift in how Vietnamese clubs structure contracts, and it flows directly from the uncertainty of the officiating system.

Effort metrics and the trap of wasted running

In the same logic, there is another category of data that I consider the most abused in Vietnamese football today: distance covered and sprint counts.

A striker covering 11 kilometres in a match sounds impressive. But if three of those kilometres are runs towards the box that end with him turning back because a teammate played square, he has burned energy without producing value. Those runs still count. They decorate the spreadsheet and leave the scoreline untouched.

I once split V.League strikers' running data into two buckets: runs oriented towards the opponent's goal while their team is in possession, and runs that produced no change in ball position. The second bucket typically accounts for thirty to forty-five per cent of total distance. For some players

Cầu thủ liên quan